Analysis of the TCR-mediated signaling
dynamics
Dissertation
zur Erlangung des akademischen Grades
doctor rerum naturalium
(Dr. rer. nat.)
genehmigt durch die Fakultät für Naturwissenschaften
der Otto-von-Guericke-Universität Magdeburg
von
M. Sc. Mateusz Pawel Poltorak
geb. am
23.03.1985 in Lublin, Polen
Gutachter: PD Dr. Luca Simeoni
eingereicht am:
16.12.2013
verteidigt am:
16.07.2014
Acknowledgements
First and foremost I would like to thank my supervisor, Dr. Luca Simeoni who was supporting me constantly throughout my whole PhD studies. From him I learned not only how to perform experiments, but also how to analyze results, judge critically obtained data, and present it in an understandable way. Without his guidance I wouldn’t be able to become a true scientist.
I am also very grateful to Prof. Burkhart Schraven not only for giving me the opportunity to work in his institute but also for continuous and intensive support of my scientific development. Despite many responsibilities he always had time to discuss, comment or solve every difficult issue.
Of course, I cannot forget to mention the great help of our technical assistants Ines Meinert and Camilla Merten, who aided me in countless experiments. I am also extremely grateful to colleagues from my group and from the institute, especially Katrin Deiser and Amelie Witte for making my work and writing this thesis as painless as possible.
Lastly, I would like to thank my friends and family for taking care of me and my life after I exit the institute.
Table of content
A. Acknowledgements 2
B. Table of content 3
C. List of figures and tables 5
D. Abstract 6
1. Introduction 7
1.1. The immune system 7
1.2. T lymphocytes 7
1.3. Molecular events occurring during T-cell activation 10
1.3.1. TCR engagement 10
1.3.2. Initiation of the TCR-mediated signaling by Src and Syk family kinases 12 1.3.3. Assembly of the LAT signalosome and activation of downstream signaling 16 1.3.4. Additional signals supporting T-cell activation 18 1.4. Regulation of T-cell activation 19 1.4.1. Feedback regulation of T-cell activation 19 1.4.2. Regulation of T-cell responses: the mode of Ras-ERK activation 25
1.5. Aims of the study 26
2. Results 28
2.1. Analysis of transient vs. sustained TCR signaling 28 2.2. Analysis of feedback regulation in transient vs. sustained T-cell activation 32 2.2.1. Transient signaling correlates with a strong activation of Src family kinases 32 2.2.2. Activation of negative regulators during transient TCR-mediated signaling 34 2.2.3. Positive feedback regulation under sustained TCR signaling 36 2.2.4. ERK-Lck feedback regulates Lck activity 38 2.3. Regulation of the Ras-ERK cascade in transient vs. sustained T-cell signaling 40 2.3.1. RasGRP1 is required for transient and sustained ERK activation 40 2.3.2. Sos1 is dispensable for transient but required for sustained ERK activation 41 2.3.3. Sos2 is dispensable for both transient and sustained ERK activation 44 2.3.4. Grb2 is dispensable for both transient and sustained ERK activation 44
3. Discussion 47
3.1. Differential regulation of proximal signaling in transient vs. sustained T-cell activation 47 3.2. Regulation of the Ras-ERK cascade differs in transient vs. sustained T-cell activation 51
4. Materials and methods 56
4.1. Ethics 56
4.2. Materials 56
4.2.1. Reagents and recipes 56
4.2.3. Antibodies 59
4.3. Methods 61
4.3.1. Human T-cell purification and culture 61
4.3.2. T-cell line culture 61
4.3.3. T-cell transfection 61
4.3.4. T-cell stimulation 62
4.3.5. Immunoblotting 62
4.3.6. Immunoprecipitation 63
4.3.7. Flow cytometric measurements 63
4.3.8. Proliferation assay 64
4.3.9. In vitro kinase assay 64
4.3.10. Statistical analysis 64
5. Used abbreviations 65
6. Bibliography 68
E. Curriculum Vitae 76
List of figures and tables
Figure 1.1. T-cell development 8
Figure 1.2. The structure of T-cell receptor (TCR/CD3) signaling complex 11 Figure 1.3. Molecular organization of TCR/CD3 complexes upon T-cell activation 13 Figure 1.4. Organization of TCR-mediated proximal signaling 15
Figure 1.5. Regulation of Lck 15
Figure 1.6. Negative and positive feedback regulation in TCR-mediated signaling 21 Figure 1.7. Structure of RasGRP1 and Sos1 23 Figure 2.1. sAbs induce transient signaling and T-cell unresponsiveness 30 Figure 2.2. iAbs trigger sustained signaling kinetics, T-cell activation and proliferation 31 Figure 2.3. Transient signaling correlates with a strong activation of Src family kinases 33 Figure 2.4. Activation of negative regulators during transient TCR-mediated signaling 35 Figure 2.5. Positive feedback regulation under sustained TCR signaling 37 Figure 2.6. ERK-Lck feedback regulates Lck activity 39 Figure 2.7. RasGRP1 is required for both transient and sustained ERK activation 42 Figure 2.8. Sos1 is dispensable for transient but required for sustained ERK activation 43 Figure 2.9. Sos2 is dispensable for both transient and sustained ERK activation 45 Figure 2.10. Grb2 is dispensable for both transient and sustained ERK activation 46 Figure 3.1. Regulation of Lck activity and signal duration 49 Figure 3.2. Regulation of the Ras-ERK cascade during transient and sustained signaling 53 Table 1.1. The role of RasGRP1, Sos1, and Grb2 in T-cell development 25 Table 3.1. Importance of RasGRP1, Sos1/2, and Grb2 in ERK activation 55
Abstract
T cells are key components of the defense system protecting our body from invading pathogens. Upon antigen recognition, T cells become activated and subsequently initiate specific cellular programs leading to proliferation, differentiation and acquisition of effector functions (e.g. cytokine production, cytotoxicity). Alterations in T-cell activation may lead to diseases such as, chronic inflammation, immunodeficiency, allergy, and cancer. Therefore in my study, I investigated how T-cell activation is regulated. In particular, I focused on the analysis of the dynamics of T-cell activation. As a model system, I used primary human T cells. Cells were stimulated to induce either transient or sustained T-cell activation. These two different activation dynamics correlate with apoptosis/unresponsiveness (anergic-like state) or proliferation, respectively. I studied how the execution of these two cellular programs is regulated at the molecular level. I found that transient signaling corresponds with strong activation of tyrosine kinases and phospho-tyrosine-dependent signaling, which induce negative feedback loops thereby terminating T-cell activation. Conversely, sustained signaling is associated with a positive feedback circuit between Lck (a crucial tyrosine kinase involved in initiation of T-cell activation) and ERK, which is required to modulate Lck activity, hence prolonging signaling.
In the second part of my work, I focused my investigation on how the dynamics of Ras-ERK activation are regulated. This cascade is critical for the specification of cellular responses in many cell types. In T cells, the Ras-ERK cascade is activated by the combined action of two guanine nucleotide exchange factors, Sos1 and RasGRP1. I found that RasGRP1 is necessary for the activation of ERK under conditions inducing both transient and sustained signaling, whereas Sos1 appears to be dispensable for transient, but required for sustained ERK signaling. In conclusion, I showed for the first time how TCR-mediated signaling dynamics are regulated in primary human T cells.
7
1. Introduction
1.1. The immune system
The natural environment is filled with a plethora of microorganisms, which may induce infections or diseases. In order to protect our body from these pathogens, we are equipped with a sophisticated network of defense mechanisms known as the immune system. The immune system can be divided into two major parts: non-specific or innate immunity and specific or adaptive immunity.
Innate immunity is the first line of defense. It responds very fast to invading pathogens, as the defenses of the innate immune system are constitutively expressed and hence they can instantly react. However, the innate immunity is not specific. This means that cells of the innate immune system recognize pathogen-associated molecular patterns (PAMPs), which are common to many pathogens. Phagocytic cells, natural killer (NK) cells, the complement system, and other secreted soluble factors constitute the innate immune system. Additionally, physical barriers like the skin and mucosa are also components of innate immunity. This type of defense is evolutionary older than adaptive immunity and it does not possess immunological memory (see below).
On the other hand, the adaptive immunity is induced upon infection and thus represents the second line of defense. It is composed of highly specialized cells, such as B and T lymphocytes, and soluble factors such as antibodies. Lymphocytes synergize together in order to protect the body from foreign pathogens. In fact, T lymphocytes are required for proper activation of B cells and antibody production. Moreover, the adaptive immunity is characterized by the presence of immunological memory. This means that after the first encounter with a pathogen the adaptive immunity develops experienced T- and B-cell subsets, which are able to rapidly respond upon a re-encounter with the same pathogen (Chaplin, 2006).
1.2. T lymphocytes
T cells are essential for adaptive immunity, as they directly recognize and respond against pathogens. Moreover, T cells are also required for the activation of other immune cells. All T cells arise from hematopoietic stem cell progenitors that are generated in the bone marrow. T-cell progenitors migrate from the bone marrow to the thymus, where they undergo a tightly regulated maturation process (Fig. 1.1). T-cell development in the thymus can be monitored by measuring the expression of surface markers defined as clusters of differentiation (CD), such as CD4 and CD8. Upon entering the thymic cortex via the post-capillary venules, lymphoid progenitors differentiate into immature T cells (thymocytes), which are characterized by the lack of CD4 and CD8 expression and are, therefore, called double
8
Figure 1.1. T-cell development. T-cell progenitors enter the thymus and undergo several stages of maturation.In the cortex they develop into DN (double negative) thymocytes and subsequently mature into DP (double positive) cells. DP cells are surveyed for their ligand binding affinity and they undergo either positive or negative selection. Cells expressing functional TCRs are then selected and commit to either the CD4 or CD8 lineage. Finally, fully mature SP (single positive) CD4+ and CD8+ T cells leave the thymus and migrate into the periphery (the figure was adopted from Germain, 2002).
9
negative (DN) thymocytes. DN thymocytes can be further subdivided by the expression of CD25 and CD44 into DN1 (CD25-, CD44+), DN2 (CD25+, CD44+), DN3 (CD25+, CD44-) and
DN4 (CD25-, CD44-) (Fig. 1.1) (Godfrey et al, 1993). Chemokines induce the migration of the
most immature thymocytes (DN1/2) from the cortex to the subcortex where they further mature into DN3 cells (Fig. 1.1).
At the DN3 stage, thymocytes express the β chain of the T-cell receptor (TCR), which pairs with an invariant pre-Tα chain and with CD3ε, CD3γ, CD3δ, and TCRζ molecules to form the pre-TCR. Signaling via the pre-TCR is ligand independent, but indispensable for further maturation of DN3 thymocytes. Cells that fail to express a functional pre-TCR are arrested at the DN3 stage and die by apoptosis. This process is one of two major checkpoints during T-cell development and is called β-selection (Germain, 2002).
The pre-TCR transduces signals inducing proliferation and further maturation to the DN4 stage. Moreover, it also initiates allelic exclusion that will ensure that only one TCR β chain will be expressed in each cell. Subsequently, DN4 cells upregulate both CD4 and CD8 molecules to become double positive (DP) thymocytes (Fig. 1.1). At this point, the β chain of the T-cell receptor forms a heterodimer with a randomly rearranged mature TCRα chain. At the end of this process, each cell bears a unique TCR. To verify the functionality of the mature TCR, a second developmental checkpoint, the so-called αβ-selection, takes place in the thymic cortex. Here, DP thymocytes are exposed to MHC (major histocompatibility complex
)
class I and II molecules complexed with self-antigens (Fig. 1.1). If a mature TCR interacts with weak/moderate affinity with self-peptide-MHC molecules, then DP cell will undergo positive selection. Positively selected cells will mature further and will be committed to either the CD4 or CD8 single positive (SP) T-cell lineages (Fig. 1.1). Conversely, if the TCR is not able to recognize self-peptide-MHC molecules, the DP thymocyte will undergo apoptosis (“death by neglect”) (Fig. 1.1). If DP thymocytes strongly react with self-peptide-MHC molecules, which indicates that they have the potential to become autoreactive, negative selection will take place. These autoagressive thymocytes will be eliminated by apoptosis, thus preventing the development of autoreactive T cells (Fig. 1.1) (Germain, 2002). These selection events are also known as central tolerance.Nevertheless, not all autoreactive T cells are deleted during negative selection. Therefore, other safety systems, called peripheral tolerance, will keep these autoaggressive T cells under control. Only a minor fraction of DP cells complete maturation and migrate to the periphery, where they circulate as naïve T cells until exposure to pathogens. Upon antigen encounter, naïve T cells will proliferate and become activated effector cells. Some of these cells will eventually differentiate into long-lasting memory T lymphocytes.
As I have mentioned before (see 1.1), memory T cells are antigen-experienced and, therefore, respond much faster to reoccurring infections, thus enhancing the adaptive
10
immune response. T-cell activation is tightly regulated in order to ensure the clearance of the pathogens without causing chronic inflammation and autoimmunity.
Mature naïve and memory T cells circulating in the periphery are divided into two major subsets, T helper (Th) and T cytotoxic (Tc) cells. These subpopulations can be distinguished
by the expression of the surface markers CD4 and CD8. Th cells express CD4, whereas
Tc cells express CD8, respectively.
Th cells function as mediators and regulators of immune responses. They orchestrate the
activation of other immune cells (e.g. B lymphocytes, macrophages etc.) by producing cytokines (i.e. interleukins). There are several subsets of Th cells, which are divided
according to their function and cytokine profile: (i) Th1 cells produce IFNγ (interferon gamma)
and activate macrophages in order to increase killing of intracellular pathogens and further support the activation of Tc cells; (ii) Th2 secrete a variety of interleukins to augment antibody
production by B cells; (iii) Th17 mainly produce IL-17 (interleukin 17) and provide
anti-microbial defense; (iv) Tregs (regulatory T cells) produce IL-10 and TGFβ (transforming growth factor beta) to suppress immune responses thereby limiting chronic inflammation and autoimmunity (Zhu et al, 2008).
Tc cells, the second T-lymphocyte subset, are responsible for the generation of cell-mediated
immunity against intracellular pathogens, such as viruses. Tc cells recognize foreign antigens
presented on MHC class I molecules, which are expressed, for example, on virus-infected cells. They kill the target cell by inducing programmed cell death (apoptosis). Apoptosis of the target cell can be induced either by the release of soluble factors such as perforin, granzymes, and granulysin, or by the engagement of the Fas receptor expressed on the target cell.
In summary, mature naïve T cells and antigen-experienced memory T cells create a versatile defense mechanism crucial for adaptive immunity.
1.3. Molecular events occurring during T-cell activation 1.3.1. TCR engagement
Given the importance of T cells within the immune system, I would like to describe the molecular events occurring during T-cell activation. First, I would like to focus my attention on the TCR, a surface receptor crucial for T-lymphocyte biology. The TCR regulates T-cell development, homeostasis, and activation. It is a heterodimer consisting of two highly variable chains, TCRα and TCRβ (or TCRγ and TCRδ in a minor T-cell population) connected by disulfide bond (Fig. 1.2). It is able to recognize peptides presented together
11
Figure 1.2. The structure of T-cell receptor (TCR/CD3) signaling complex. The TCR consists of two chainsTCRα and TCRβ (shown in blue) containing constant (C) and variable (V) regions. The TCR/CD3 complex is formed from CD3δε and CD3γε heterodimers (shown in green), and two TCRζ chains (shown in violet). The ITAMs (immunoreceptor tyrosine-based activation motifs) required for signal initiation are indicated in red.
12
with
MHC class I and II molecules. The TCRα and TCRβ chains of the TCR are composed of constant (C) and variable (V) regions (Fig. 1.2). The constant part is required for anchoring of the TCR to the plasma membrane, whereas the variable part contains the antigen-binding site (Fig. 1.2). V regions are generated by random joining of gene segments, thus creating the unique antigen specificity of the TCR. The TCRαβ heterodimer is associated with the CD3 and TCRζ molecules required for signal transduction. The full TCR/CD3 complex consists of a TCRαβ heterodimer, CD3εδ and CD3εγ heterodimers, and a TCRζζ homodimer (Fig. 1.2). Signaling downstream of the TCR/CD3 complex depends on the phosphorylation of distinct tyrosine motifs called immunoreceptor tyrosine-based activation motifs (ITAMs) located within the CD3 and TCRζ chains (Fig. 1.2). ITAMs include two characteristic amino acid sequences (YxxL or YxxI) separated from each other by 6 to 8 amino acids. The TCR/CD3 complex contains 10 ITAMs.T-cell activation is initiated upon the binding of the TCR to peptide antigens presented by MHC molecules expressed on antigen presenting cells (APCs). However, despite intense investigations, it is not yet fully understood how TCR-mediated signaling is initiated and several models have been hypothesized. The widely accepted hypothesis, called the segregation model, postulates that signals are triggered upon spatial reorganization of TCR/CD3 complexes and effector molecules (Fig. 1.3) (Davis et al, 2006). According to this model, TCR engagement leads to TCR oligomerization and to the formation of signaling microclusters also including crucial effector molecules such as Lck (lymphocyte-specific protein tyrosine kinase), but not negative regulators such as the phosphatase CD45 and the tyrosine kinase Csk (c-Src tyrosine kinase) (Fig. 1.3) (Torgersen et al, 2001; Choudhuri et al, 2010; Borger et al, 2013; Rossy et al, 2013). Thus, according to this model, signaling is initiated upon the segregation of positive and negative regulatory molecules.
1.3.2. Initiation of the TCR-mediated signaling by Src and Syk family kinases
The TCR has no intrinsic catalytic activity and therefore it is closely associated with tyrosine kinases belonging to the Src and Syk family. Src (sarcoma tyrosine kinase) family kinases (SFKs) such as Lck and Fyn (feline yes-related protein) phosphorylate the ITAMs within the CD3 and TCRζ chains. It has been proposed that both Lck and Fyn are constitutively active (40% - 50% of the total pool) in T cells and maintain the basal level of TCR phosphorylation (Nika et al, 2010; Brownlie et al, 2013). This is necessary to provide tonic signaling required for T-cell survival (Seddon et al, 2002). In T cells, Lck can be associated with the CD4 or CD8 co-receptors expressed on the surface of Th or Tc cells, respectively. Upon TCR
engagement, Lck is brought into close proximity of the ITAM chains in both a co-receptor-dependent and -inco-receptor-dependent manner (Artyomov et al, 2010). According to the segregation
13
Figure 1.3. Molecular organization of TCR/CD3 complexes upon T-cell activation. (a) In resting T cells,TCR/CD3 complexes, effector molecules (e.g. Lck), and negative regulators (e.g. Csk, CD45) are randomly distributed throughout the plasma membrane. (b) Upon T-cell activation, TCR complexes oligomerize with effector molecules to form microclusters, whereas negative regulators are excluded from this clustering zone.
14
model discussed above (see 1.3.1), Lck and TCR/CD3 complexes are then sequestered from inhibitory molecules into microclusters (Fig. 1.3), where Lck will phosphorylate the ITAMs, thus initiating TCR-mediated signaling (Fig. 1.4).
Lck is the apical tyrosine kinase in the signaling cascade and the mechanisms regulating Lck activation have been under intense investigation for more than 20 years. However, how Lck is activated upon TCR triggering is still not well-understood. According to the model proposed by Nika et al., and others, signaling is initiated upon translocation of the active Lck pool to the engaged TCR without the need of de novo Lck activation (Paster et al, 2009; Nika et al, 2010). However, a very recent study from our institute has challenged this model and proposed that a fraction of Lck is indeed activated at the triggered TCR (Stirnweiss et al, 2013).
Lck contains an N-terminal unique region (with two palmitoylation and one myristoylation sites required for lipid raft anchoring), an SH3 (Src homology 3) domain (necessary for interaction with other proteins via their proline-rich regions) followed by an SH2 domain (which binds phosphorylated tyrosine residues) and the kinase domain followed by a C-terminal tail (Fig. 1.5a). In addition, Lck possesses two critical tyrosine residues, Y505
located in the C-terminus and Y394 located within the activatory loop of the kinase domain,
which control Lck activation (Fig. 1.5a). Phosphorylation of Lck on Y505 results in the binding
of the C-terminal tail to the Lck-SH2 domain thereby, generating a “closed” conformation and an inactive enzyme (Fig. 1.5b) (Xu et al, 1999). The phosphorylation of Y505 is mediated by Csk, which is a master negative regulator of TCR-mediated signaling (Schoenborn et al, 2011). Conversely, the phosphatase CD45 is known to dephosphorylate Lck at the C-terminal tyrosine, leading to the so-called “primed” non-phosphorylated Lck (Fig. 1.5b) (Hermiston et al, 2003; Salmond et al, 2009). At this point, Lck can cluster and
trans-phosphorylate on Y394 leading to the “opened” conformation, which corresponds to an
active enzyme (Fig. 1.5b). It has been proposed that the “opened” conformation can be reverted by the phosphatases PTPN22 (protein tyrosine phosphatase non-receptor type 22) bound to Csk, SHP1 (SH2 domain-containing phosphatase 1), or CD45 (Cloutier et al, 1999; Salmond et al, 2009). Finally, “opened” Lck can be further phosphorylated on Y505 resulting in
double phosphorylated, active form (Nika et al, 2010). Only “opened” and active Lck is capable to phosphorylate ITAMs and therefore initiate signaling.
15
Figure 1.4. Organization of TCR-mediated proximal signaling. After T-cell activation Lck phosphorylates theITAMs of the TCR/CD3 complex, allowing recruitment and Lck-mediated activation of ZAP70. Subsequently, activated ZAP70 phosphorylates LAT and SLP76 adaptor proteins, thus facilitating formation of the LAT signalosome. The assembled LAT signalosome consists of LAT, Grb2, Gads, SLP76, ADAP, PLCγ1, Itk, Nck, and Vav1. Here, Itk is able to phosphorylate PLCγ1 which subsequently hydrolyzes PIP2 to DAG and IP3, second
messengers required for the activation of PKC-, Ras-, and Ca++-mediated downstream pathways necessary for activation of the transcription factors AP1, NFκB, and NFAT, respectively. Activated AP1, NFκB, and NFAT drive the synthesis of the cytokine IL-2 to further support T-cell activation and proliferation. Phosphorylation of crucial molecules is indicated by red dots.
Figure 1.5. Regulation of Lck. (a) Schematic representation of the Lck protein. Blue squares represent domains
and red circles indicate possible phosphorylation sites with the corresponding amino acid indicated below. The structure of Lck is as follows: unique domain (UD), Src homology domain 3 (SH3), Src homology domain 2 (SH2), kinase domain (also known as SH1), and tail region. (b) Conformational changes of Lck representing “closed”, “primed” or “opened” form (figure 1.5b was modified from Acuto et al, 2008).
16
TCR-mediated signaling is further propagated upon recruitment of ZAP70 (zeta-chain-associated protein kinase 70 kDa), a member of the Syk family, to the phosphorylated ITAMs (Fig. 1.4) (Isakov et al, 1995). Binding of its tandem SH2 domains to the ITAMs unlocks ZAP70, which is subsequently phosphorylated on Y315 and Y319 by Lck. Further
auto-phosphorylation on Y493 completes ZAP70 activation (Pelosi et al, 1999; Deindl et al, 2007; Acuto et al, 2008; Yan et al, 2013). Catalytically active ZAP70 is necessary for the phosphorylation of two adaptor proteins, LAT (linker of activated T cells) and SLP76 (SH2 domain-containing leukocyte protein of 76 kDa), which constitute the center of an anchoring platform (LAT signalosome), which recruits additional cytosolic signaling molecules that are required to propagate signaling to the nucleus (Fig. 1.4) (Bubeck-Wardenburg et al, 1996; Paz et al, 2001).
1.3.3. Assembly of the LAT signalosome and activation of downstream signaling
LAT belongs to the transmembrane adaptor protein (TRAP) family and has nine tyrosine residues, which are phosphorylated upon TCR triggering (Fuller et al, 2011; Balagopalan et al, 2010). Phosphorylated LAT binds effector proteins such as PLCγ1 (phospholipase C gamma 1), the p85 subunit of PI3K (phosphoinositide 3-kinase), and the cytosolic adaptors - Grb2 (growth factor receptor-bound protein 2) and Gads (Grb2-related adapter protein downstream of Shc) (Fig. 1.4). Gads is constitutively associated with SLP76 (Liu et al, 1999). Upon recruitment of Gads to LAT, SLP76 interacts with PLCγ1, the proto-oncogene Vav1, the Tec family kinase Itk (IL-2-inducible T-cell kinase), and the adaptor proteins - Nck (non-catalytic region of tyrosine kinase adaptor protein) and ADAP (adhesion and degranulation promoting adapter protein) (Fig. 1.4). Thus, by recruiting several cytosolic effector molecules, LAT and SLP76 coordinate the activation of different cellular signaling pathways (e.g. Ca++,
PKC, Ras), which will ultimately culminate in gene expression, cytoskeletal reorganization, and T-cell activation (Smith-Garvin et al, 2009).
The activation of the phospholipase PLCγ1 is of particular importance during T-cell activation. Upon binding to LAT and SLP76, PLCγ1 is phosphorylated on Y783 and hence
activated by Itk. Subsequently, PLCγ1 hydrolyzes the membrane lipid PIP2
(phosphatidylinositol-4,5-bisphosphate), producing the second messengers IP3 (inositol
1,4,5-trisphosphate) and diacylglycerol (DAG) (Fig. 1.4). IP3 activates Ca++-dependent
signaling, whereas DAG will initiate two major intracellular signaling pathways involving protein kinase C (PKC) and Ras (rat sarcoma) (Fig. 1.4) (Smith-Garvin et al, 2009). These pathways lead to the activation of three transcription factors – NFAT (nuclear factor of activated T-cells), NFκB (nuclear factor kappa light chain enhancer of activated B cells), and AP1 (activator protein 1), which are required for IL-2 synthesis (Fig. 1.4). IL-2 is an important cytokine for activated T cells, which enhances T-cell activation and proliferation. Therefore,
17
given the importance of NFAT, NFκB, and AP1, I will briefly describe the molecular events leading to their activation.
Ca++-NFAT
PLCγ1-generated IP3 activates IP3 receptors on the endoplasmic reticulum (ER) and leads to
the release of Ca++ from cytoplasmic stores. Depletion of intracellular calcium opens calcium release activated channels (CRAC) on the plasma membrane, thus allowing the entry of extracellular Ca++. Rise of intracellular Ca++ is sensed by calmodulin (CaM), which binds and
activates other proteins such as, calmodulin kinase (CaMK) and the calmodulin-dependent phosphatase calcineurin. Calcineurin dephosphorylates members of the NFAT transcription factor family, thus allowing their translocation into the nucleus to activate gene transcription (Smith-Garvin et al, 2009, Robert et al, 2011).
PKCθ-NFκB
The NFκB pathway is initiated by the generation of DAG at the plasma membrane. DAG production results in the membrane localization and activation of PKCθ. PKCθ is then able to phosphorylate CARMA1 (CARD-containing MAGUK protein 1), a member of a trimolecular CBM complex (CARMA1, Bcl10, and MALT1). Upon phosphorylation CARMA1 oligomerizes and associates with Bcl10 (B-cell lymphoma 10), and MALT1 (mucosa-associated lymphoid tissue lymphoma translocation protein 1). The CBM complex contributes to the degradation of the regulatory subunit of IKK (IκB kinase), thus unlocking the IKK catalytic site. Activated IKK phosphorylates IκB (inhibitor of kappa B) which keeps NFκB inactive. Phosphorylation of IκB leads to its degradation and release of NFκB. Ultimately, free NFκB molecules enter the nucleus and initiate gene transcription (Smith-Garvin et al, 2009).
Ras-ERK-AP1
Activation of the AP1 transcription complex, consisting of two proto-oncogenes Jun and Fos, depends on the Ras-ERK pathway. Ras is a small guanine nucleotide-binding protein that hydrolyzes GTP into GDP (GTPase). Ras is active in the GTP-bound state and inactive when loaded with GDP. Transition from the inactive to the active state is mediated by GEFs (guanine nucleotide exchange factors), which facilitate the release of GDP and promote the binding of GTP. During T-cell activation, GEFs are induced upon the assembly of the LAT signalosome (see 1.4.1).
In activated T cells, RasGTP (active Ras) interacts with Raf (rat fibrosarcoma, MAP3K), a MAPK (mitogen-activated protein kinase) through its Ras binding domain (RBD). This association is necessary to unlock autoinhibited Raf from its “closed” conformation. “Opened” Raf can dimerize and trans-phosphorylate to achieve full activation. In turn, activated Raf phosphorylates and activates the dual specificity kinases MEK1/2 (mitogen-activated protein kinase kinase 1/2). MEKs are required for the phosphorylation of the crucial serine/threonine-specific protein kinases ERK1/2 (extracellular signal-regulated kinase 1/2). Activated ERK is
18
essential for the generation of the AP1 transcription complex. ERK1/2 can modulate gene transcription directly (by phosphorylating and stabilizing Jun and Fos) or indirectly (by phosphorylating RSK [ribosomal s6 kinase] and ELK [ETS domain-containing protein] transcription activators). The phosphorylated AP1 complex controls the transcription of several genes and drives the expression of CD69, which is commonly used as a marker of T-cell activation (Smith-Garvin et al, 2009).
1.3.4. Additional signals supporting T-cell activation
CD28-mediated co-stimulation
In order to induce full T-cell activation, co-stimulatory signals are also required. In fact, ligation of the TCR alone (defined as signal 1) does not induce T-cell activation, but results in a non-responsive state called anergy. In addition to the TCR, co-stimulatory receptors deliver signals (called signal 2) during T-cell activation. Ligands for these receptors are expressed on APCs, such as dendritic cells (DCs), macrophages, and B cells. Among the co-stimulatory receptors in T cells, CD28 is one of the most studied. CD28 interacts with CD80 (B7-1) and CD86 (B7-2) molecules expressed on APCs. CD28-mediated co-stimulation synergies with T-cell receptor signals and promotes survival, clonal expansion, and differentiation (Rudd et al, 2009).
Similarly to the TCR, CD28 lacks intrinsic catalytic activity and signals by recruiting effector proteins to its cytoplasmic tail. Engagement of CD28 by its ligands leads to Lck and Fyn-mediated tyrosine phosphorylation of tyrosine-based signaling motifs (TBSMs) located within the CD28 cytosolic tail. The phosphorylation of the YMNM motif promotes the association of CD28 with the p85 regulatory subunit of PI3K and Grb2 (Okkenhaung et al, 1998). Recruited p85 binds to p110, the catalytic subunit of PI3K, which converts the phospholipid PIP2 into
PIP3. In turn,PIP3 serves as a docking site for pleckstrin homology (PH) domain-containing
proteins including PDK1 (phosphoinositide-dependent protein kinase 1) and its substrate PKB/Akt (protein kinase B/Ak thymoma). Recruitment of these molecules to the plasma membrane activates signaling pathways involved in cell metabolism and survival (Rudd et al, 2009; Smith-Garvin et al, 2009).
IL-2-mediated signaling
In addition to TCR- and CD28-mediated signaling (signal 1 and 2), cytokines such as IL-2 (signal 3) also play a role in T-cell activation and proliferation. TCR- and CD28-mediated transcriptional activation leads to IL-2 production and to the upregulation of the IL-2R (IL-2 receptor). IL-2-IL-2R interaction provides a mitogenic signal leading to clonal expansion of activated T cells (Malek, 2008). The high affinity form of the IL-2 receptor is composed of IL-2Rα (CD25), IL-2Rβ (CD122), and the common γ chain (CD132). The latter is also shared with other cytokine receptors. The IL-2Rα and β chains bind IL-2. This leads to the activation
19
of Janus kinase 3 (JAK3). Subsequently, the IL-2Rβ chain is phosphorylated by active JAK3, thus enabling the recruitment of JAK1 and Shc (SH2 domain-containing transforming protein) to the receptor. At the activated IL-2R, triple phosphorylated Shc serves as a platform for the binding of Grb2/Sos1 complex required for the activation of the Ras-ERK cascade (see 1.4.1) (Gu et al, 2000). On the other hand, activated JAK1/3 phosphorylate STAT (signal transducer and activator of transcription) proteins, which dimerize and translocate to the nucleus where they can regulate cell growth, survival and differentiation (Vainchenker et al, 2013). The IL-2R can also recruit the p85 subunit of PI3K, which will trigger PKB/Akt activation, thus further promoting T-cell survival. Interestingly, the IL-2R and TCR signaling networks are interconnected. The cross-talk between these two signaling pathways has been recently described (Beyer et al, 2011). This study suggests that for adequate T-cell activation some effector proteins, such as ERK, have to receive all three activatory signals.
1.4. Regulation of T-cell activation
1.4.1. Feedback regulation of T-cell activation
T-cell activation is regulated by the interplay between upstream activators, such as apical tyrosine kinases, and effector proteins downstream. The interactions between these signaling molecules are defined as positive and negative feedback loops. Feedback regulation determines whether agonist-induced activation will be translated into transient or sustained T-cell signaling, or whether it will be terminated. Below I would like to describe several important examples.
Negative feedback regulation
Negative feedbacks circuits are crucial during T-cell activation as they are responsible for fine-tuning and termination of the signal. Negative regulatory mechanisms oppose rapid induction and amplification of biochemical events. They allow signals to be controlled or stopped and guarantee the appropriate response to perturbations. Phosphorylation and dephosphorylation are the main means of dampening signal propagation, although other modifications, such as ubiquitinylation, may also contribute (Acuto et al, 2008).
Very strict and elaborate feedback regulation adjusts proximal signaling. For example, the Lck-SHP1 negative feedback loop controls Lck activity. As I have mentioned previously (see 1.3.2), SHP1 is one of the phosphatases involved in the inhibition of Lck. Upon T-cell activation, Lck is able to phosphorylate SHP1 on several tyrosine residues including Y564
which lies within the consensus sequence for binding of the Lck SH2 domain (Stefanova et al, 2003). Phosphorylated SHP1 can deactivate Lck directly by dephosphorylating the activatory Y394 residue (Fig. 1.5b). SHP1 can also counteract Lck action by
20
dephosphorylating Lck substrates or other downstream signaling molecules, such as TCRζ, ZAP70, Vav1, Grb2, and SLP76. (Fig. 1.6a) (Lorenz et al, 1994; Plas et al, 1996; Acuto et al, 2008; Methi et al, 2005).
Another negative feedback loop regulating Lck, and also Fyn, involves the adaptor protein PAG (phosphoprotein associated with GEMs) and Csk (Brdicka et al, 2000; Smida et al, 2007). In resting T cells, membrane-bound PAG is constitutively phosphorylated by Fyn. Phosphorylated PAG binds Csk and localizes it at the plasma membrane, in the proximity of SFKs. There, Csk is able to phosphorylate the inhibitory tyrosines of Lck and Fyn, Y505 and
Y529, respectively (Fig. 1.6a). Upon T-cell activation, PAG is dephosphorylated, likely by
CD45, and Csk is released into the cytoplasm, where it no longer inhibits tyrosine kinases (Brdicka et al, 2000).
Other components of proximal signaling are regulated by a negative feedback loop mediated by the SHIP1-Dok2 module (Fig. 1.6a) (Acuto et al, 2008). Upon TCR stimulation, SHIP1 (SH2 domain-containing inositol-5-phosphatase) and Dok2 (docking protein 2) are phosphorylated by Tec family tyrosine kinases and form a complex with Grb2. SHIP1 dephosphorylates PIP3 into PIP2, thus interfering with the recruitment of PH
domain-containing proteins, such as PKB/Akt or PDK1, to the plasma membrane (Smith-Garvin et al, 2009; Acuto et al, 2008). In addition, Dok2 negatively regulates signaling by recruiting Csk or RasGAP (Ras GTPase activating protein) – an inhibitor of the Ras-ERK pathway (Schoenborn et al, 2011; Acuto et al, 2008). Furthermore, it has been proposed that Dok2, together with Dok1, may also compete with ZAP70 for binding to the phosphorylated ITAMs or interfere with the assembly of the LAT signalosome, thus negatively regulating T-cell activation in multiple ways (Fig. 1.6a) (Dong et al, 2006; Yasuda et al, 2007).
Ubiquitinylation of components of the TCR/CD3 complex and also other effector molecules plays an additional role in the inhibition of TCR-mediated signaling. It has been shown that members of the CBL (casitas B lineage lymphoma) family (e.g. cCbl, an E3 ubiquitin ligase) are involved in the ubiquitinylation and subsequent degradation of ZAP70, TCRζ chains, and potentially other components of the TCR/CD3 complex. In activated T cells, cCbl is brought into close proximity of the ζ chains upon its binding to phosphorylated ZAP70. Here, cCbl may promote ubiquitinylation of its targets (Fig. 1.6a) (Wang et al, 2001; Naramura et al, 2002; Wang et al, 2008). This process is also thought to be part of the mechanism regulating TCR expression.
Positive feedback regulation
In contrast to negative regulation, positive feedbacks promote signaling. Positive circuits are crucial for the prolongation and/or the amplification of the initial signal without continuous presence of the original stimulus. Extending and propagating signals is necessary for
21
Figure 1.6. Negative and positive feedback regulation in TCR-mediated signaling. (a) The following negativeregulators are involved in the orchestration of T-cell signaling, PAG-bound Csk (a master tyrosine kinase regulator), SHP1 (a phosphatase able to dephosphorylate Lck, ZAP70, and other signaling molecules), cCbl (an E3 ubiquitin ligase capable of ubiquitinylating TCRζ, ZAP70, and potentially other components of TCR complex), and SHIP1-Dok2 complex (responsible for deactivation of Ras and interferening with ITAMs phosphorylation), which is phosphorylated by Tec kinases (TK). (b) Actions of some negative regulators can be counteracted by positive feedbacks, such as the ERK-Lck feedback loop, which prevents Lck from SHP1-mediated dephosphorylation or Ras-Sos1 feedback necessary to amplify activation of Ras-ERK cascade (the figure was modified from Poltorak et al, 2013).
22
facilitation of particular molecular processes, for example stabilization of transcription factors. The interplay between positive and negative regulation shapes T-cell responses and consequent functional choices.
As a model molecule, Lck is regulated by both positive and negative feedbacks. A positive feedback between ERK1/2 and Lck has been proposed to block the interaction with SHP1 (see above) (Stefanova et al, 2003; Dong et al, 2010). It has been shown that Lck can be phosphorylated by activated ERK1/2 on serine 59 (S59) (Watts et al, 1993; Winkler et al,
1993; August et al, 1996). Phosphorylation of this site leads to a conformational change in Lck, which affects the binding capacity of the Lck SH2 domain and prevents the interaction of Lck with SHP1, thus impeding SHP1-mediated dephosphorylation and hence inactivation of Lck (Fig. 1.6b) (Joung et al, 1995; Stefanova et al, 2003).
In addition, activation of the Ras-ERK cascade is regulated by an unusual interplay of Ras activators involving a positive feedback circuit. T cells express GEFs belonging to two different families, RasGRP1/4 (Ras guanyl-releasing protein 1/4) and Sos1/2 (son of sevenless 1/2) (Kortum et al, 2013; Stone, 2011). Although the function of RasGRP4 is not yet fully understood, it is very well-established that RasGRP1 is the major Ras activator in T cells (Genot et al, 2000; Smith-Garvin et al, 2009). Studies from RasGRP1-/- mice and
T-cell lines demonstrated the importance of RasGRP1 for the activation of ERK in both mature and immature T cells (Roose et al, 2005; Roose et al, 2007; Priatel et al, 2010; Kortum et al, 2012). Moreover, defects in the expression of RasGRP1 in humans are contributing factors to autoimmune diseases, such as systemic lupus erythematosus (SLE) (Yasuda et al, 2007; Stone, 2011). RasGRP1 possesses a catalytic domain required for its GEF function composed of a REM (Ras exchange motif) box and a CDC25 (cell division cycle 25) box (Fig. 1.7a). Additionally, RasGRP1 possesses two calcium-binding elements called EF hands, a C1 domain for DAG-binding, and a unique tail (Fig. 1.7a) (Ebinu et al, 1998; Stone, 2011). Upon T-cell activation RasGRP1 is recruited to the plasma membrane via its C1 domain and the unique tail. At the plasma membrane, RasGRP1 is activated by PKCθ-mediated phosphorylation of T184 (Carrasco et al 2004; Roose et al, 2005; Fuller et al, 2012). Thus, the activation of RasGRP1 depends on DAG and hence on PLCγ1 activity. The second Ras activator, Sos, was discovered in Drosophila melanogaster, where it is essential for normal eye development (Bonfini et al, 1992). In human T cells, two homologues are expressed, Sos1 and Sos2 (Chardin et al, 1994). Mutations in the Sos1 gene have been recently reported in Noonan syndrome, which is a RASopathy – a developmental disorder caused by alterations in genes related to Ras-MAPK pathways (Pierre et al, 2011). In contrast to RasGRP1, Sos1 is constitutively bound to Grb2, an adaptor which is recruited to phosphorylated LAT (see 1.3.3). Thus, Grb2 is required to bring
23
Figure 1.7. Structure of RasGRP1 and Sos1. (a) Schematic representation of the RasGRP1 protein. Greensquares represent domains. The structure of RasGRP1 is as follows: catalytic domain-containing a REM (Ras exchange motif) box and a CDC25 (cell division cycle 25) box, EF hands, and a C1 domain. (b) Schematic representation of the Sos1 protein. Orange squares represent domains. A unique allosteric pocket required for RasGTP binding is indicated. The structure of Sos1 is as follows: Dbl homology (DH) domain, pleckstrin homology (PH) domain, catalytic domain containing REM box, CDC25 box and unique allosteric pocket, and proline-rich region (PRR).
24
Sos to the plasma membrane. Sos1 has a Dbl homology (DH) domain, a PH domain followed by a RasGRP1-like catalytic domain with a unique allosteric pocket, and a proline-rich region (PRR) necessary for interaction with the SH3 domain of Grb2 (Fig. 1.7b) (Margarit et al, 2003; Pierre et al, 2011). Upon TCR stimulation, the Grb2/Sos1 complex is recruited to phosphorylated LAT where it can facilitate the activation of Ras (Genot et al, 2000; Smith-Garvin et al, 2009). It is believed that the intrinsic GEF activity of Sos1 can be greatly enhanced upon the loading of active RasGTP in its unique allosteric pocket (Fig. 1.7b) (Margarit et al, 2003; Roose et al, 2007). Moreover, it has been proposed that membrane-recruited Sos1 is in an autoinhibited conformation and priming of Sos1 initially depends on RasGTP generated exclusively by RasGRP1 (Margarit et al, 2003; Sondermann et al, 2004). According to a recently proposed model, it is believed that Ras activation in T cells is orchestrated by the coordinated action of both RasGRP1 and Sos1 (Roose et al, 2007; Das et al, 2009). Initial TCR triggering leads to Ras activation exclusively via RasGRP1. Subsequently, RasGRP1-generated RasGTP primes Sos1, thus enhancing its enzymatic activity up to 80-fold. This interaction creates a positive RasGTP-Sos1 loop and strongly increases the levels of active Ras (Fig. 1.6b) (Roose et al, 2007; Das et al, 2009). It has been proposed that RasGTP generated by RasGRP1 alone regulates T-cell differentiation (positive selection), whereas the high amount of RasGTP generated by both RasGRP1 and Sos is believed to take part in the regulation of apoptosis in developing thymocytes (negative selection) (Priatel et al, 2002; Prasad et al, 2009). However, recent studies suggest that at specific stages of T-cell development, RasGRP1 and Sos1 can activate Ras independently (Kortum et al, 2011; Kortum et al, 2012). For example, RasGRP1-/- mice display only a mild defect in β-selection, but a severe block in positive selection (Table 1.1) (Kortum et al, 2012). On the other hand, Sos1 conditional knock-out mice show a strong defect in β-selection, but normal positive and negative selection (Table 1.1) (Kortum et al, 2011). Therefore these studies indicate that Sos1 can influence T-cell development during particular stages of thymopoiesis independently of RasGRP1 (e g. during β-selection).
25
Mouse model Developmental effect
RasGRP1-/- 20% reduction in thymocyte cellularity
Mild defect in β-selection Block in positive selection Normal negative selection
Sos1-/- 50% reduction in thymocyte cellularity
Proliferative defect at the β-selection checkpoint Normal positive and negative selection
RasGRP1/Sos1-/- Up to 90% reduction in thymocyte cellularity Block in β-selection
Block in positive and negative selection Grb2-/- Defective positive and negative selection
Table 1.1. The role of RasGRP1, Sos1, and Grb2 in T-cell development (the table
modified from Kortum et al, 2013)
1.4.2. Regulation of T-cell responses: the mode of Ras-ERK activation
The magnitude/duration of Ras-ERK activation controls cellular responses. In the well-established model system for neuronal differentiation based on the PC12 cell line, it has been shown that moderate and sustained ERK phosphorylation upon NGF (nerve growth factor) treatment causes cell differentiation, whereas strong and transient ERK activation upon EGF (epidermal growth factor) stimulation induces proliferation in the same cells (Santos et al, 2007). In other cell types such as fibroblasts, it has been demonstrated that sustained ERK activation induced by mitogens, such as α-thrombin, leads to cell cycle re-entry, whereas transient signaling triggered by synthetic agonists, such as TMP (thrombin mimicking peptide), results in quiescence (Vouret-Craviari et al, 1993; Murphy et al, 2002). Thus, data from different cell types indicate that receptor stimulation at the plasma membrane is translated into quantitatively and/or qualitatively different activation kinetics of the Ras-ERK module to activate different cellular programs.
In T cells, it has been postulated that the magnitude of ERK activation is important for the regulation of thymic development. For example, strong ERK activation correlates with apoptosis (negative selection), whereas weak ERK activity is associated with differentiation (positive selection) (Daniels et al, 2006). Additionally, it has been shown that also in mature mouse T cells the magnitude of ERK activation is important for the regulation of T-cell responses. For example, strong ERK signal in CD8+ T cells leads to apoptosis, in contrast to
26
The duration of the ERK signal is another important factor for the generation of cellular responses. A dynamic behavior of ERK activity seems to exist in murine thymocytes. Transient ERK activation induces negative selection, whereas sustained ERK activity leads to positive selection of immature T cells (McNeil et al, 2005; Daniels et al, 2006). In mature T cells, induction of transient signals corresponds with unresponsiveness or apoptosis, in contrast to sustained ERK signaling, which results in activation and proliferation (Berg et al, 1998; Wang et al, 2008).
It has been proposed that differences in the magnitude and the duration of the Ras-ERK signaling are regulated by different mechanisms including feedback loops (see 1.4.1). In PC12 cells a strong and transient signal results in activation of a negative feedback circuit between ERK and, most likely, Sos, whereas moderate and sustained ERK phosphorylation is regulated by PKC-mediated ERK-Raf positive feedback loop (Santos et al, 2007). In cytotoxic mouse T lymphocytes transient signaling is mediated by classical PKC isoforms, whereas novel PKCs are involved in prolonged signals (Puente et al, 2006). Additionally, it has been shown in mouse thymocytes that the compartmentalization of Ras activators (Grb2/Sos1 complex and RasGRP1), as well as Ras and ERK themselves may contribute to the activation kinetics. Localization analysis of these molecules indicated that during transient activation, signaling occurs exclusively at the plasma membrane. However, when sustained signaling is induced, the activated proteins localize in cytoplasmic vesicles (Daniels et al, 2006; Wang et al, 2008).
Despite the fact that the regulation of Ras-ERK activation has been well-studied in some cell systems, how the magnitude/duration of the Ras-ERK signal is regulated in primary human T cells is still poorly understood.
In summary, in order to activate the appropriate cellular program and to induce an efficient immune response, T cells are equipped with a multitude of regulatory mechanisms, which serve to integrate, fine-tune, and terminate signals triggered at the plasma membrane.
1.5. Aims of the study
Altered T-cell activation is the basis for many human diseases, such as chronic inflammation, immunodeficiency, allergy, and cancer. In my study, I focused on how TCR-mediated signaling is regulated on the molecular level. For my experiments, I used primary human T cells because of their relevance in human disease. I addressed the following scientific questions:
How are the activation dynamics regulated in T cells?
In the first part of my work, I analyzed how transient vs. sustained TCR signaling is regulated. Transient signaling correlates with apoptosis/unresponsiveness (anergic-like
27
state), whereas sustained signaling is associated with T-cell activation and proliferation. I found that transient signaling is regulated by negative regulatory loops involving inhibitory molecules such as Dok2 and cCbl, whereas sustained signaling is triggered by a positive regulatory feedback involving the ERK-mediated phosphorylation of Lck.
How is Ras-ERK activation regulated in T cells?
In the second part of my work, I focused my attention on the contribution of RasGRP1 and Sos1 to Ras-ERK activation. Using RNAi (RNA interference), I demonstrated that RasGRP1 is a crucial activator of Ras-ERK in primary human T cells. Conversely to RasGRP1, Sos1 contributes only to Ras-ERK activation during sustained signaling. Moreover, I found that Sos2, a homologue of Sos1, and Grb2, an adaptor molecule associated with Sos1, appear to be dispensable for the ERK activation upon TCR-mediated stimulation in human primary T cells.
28
2. Results
In the results section, I characterized stimulation methods to induce transient or sustained TCR-mediated signaling, investigated the activation dynamics of key signaling molecules upon TCR triggering and analyzed the regulation of Ras-ERK activation.
2.1. Analysis of transient vs. sustained TCR signaling
In order to study how T cells are activated upon physiological stimulation in living organisms, the in vitro experimental setup has to mimic as closely as possible the in vivo conditions. During the past years, different methods have been developed to stimulate T cells in vitro, which have allowed an extensive analysis of the biochemical events occurring during T-cell activation. A variety of studies have employed mouse T cells expressing a transgenic TCR. The advantage of this system is that T cells can be stimulated with physiological ligands (e.g. peptide-MHC complexes specific for the transgenic TCR). However, the mouse immune system is not fully comparable with the human immune system. In fact, a very recent study has analyzed the profiles of genes upregulated in response to inflammatory stress in humans and mice and has come to the conclusion that mouse models are poorly suited to study human inflammatory diseases (Seok et al, 2013). Therefore, in order to better understand the molecular mechanisms underlying human diseases, I decided to employ human T cells for my studies. Unfortunately, with the exception of memory T cells, which can be re-stimulated
in vitro with specific antigen (e. g. tetanus) (Cellerai et al, 2007), there are no physiological
ligands available to stimulate a sufficient number of naïve human T cells in vitro for biochemical studies.
Therefore, I took advantage of two available stimulation systems to activate human peripheral T cells in vitro. I used antibodies applied in solution (sAbs) and antibodies immobilized on microbeads (iAbs). Both systems are based on antibodies directed against the TCR/CD3 complex and co-stimulatory molecules. Anti-CD3 antibodies recognize epitopes in the extracellular part of the CD3ε chains and induce the aggregation of at least two TCR/CD3 complexes (a process called crosslinking). Antibodies against the CD28 co-stimulatory molecule and the CD4 co-receptor utilize the same principle. In my studies, I used monoclonal IgG antibodies (mAbs) against human CD3ε (clones OKT3, UCHT1), CD28 (clone CD28.2), and CD4 (clone OKT4), which were biotinylated. Biotinylation allows linkage of multiple antibodies upon the addition of streptavidin, thus enhancing TCR/CD3 crosslinking and T-cell stimulation,
One of the most important results of my studies is that sAbs and iAbs induce completely different signaling dynamics and T-cell functional outcomes (Arndt et al, 2013; Poltorak et al, 2013). Therefore, these stimuli have allowed me to study the molecular mechanisms
29
regulating signaling kinetics and T-cell responses. The major characteristics of transient and sustained signaling triggered by sAbs or iAbs, respectively, are presented below.
Transient signaling induced by sAbs
Human peripheral T cells were left unstimulated or stimulated with sAbs for the indicated time periods (Fig. 2.1). The Western blot analysis presented in figure 2.1 clearly show that the maximal signal intensity induced by sAbs peaks at 2 - 5 min and then rapidly declines to the basal levels 15 to 30 min post-stimulation (Fig. 2.1a). These transient phosphorylation kinetics are observed for molecules involved in both proximal (ZAP, LAT, PLCγ1) and more distal (Akt, Raf, MEK, ERK, RSK) TCR signaling (Fig 2.1a and Poltorak et al, 2013).
It has been previously shown that transient TCR-mediated signaling does not induce activation and differentiation of mouse CD8+ T cells (Berg et al, 1998; Wang et al, 2008). To test if sAbs also do not induce T-cell activation in human primary T cells, I measured the expression of CD69 and CD25, two well-known activation markers. CD69 is expressed in response to the ERK-dependent activation of the transcription factor AP1 and CD25 upregulation occurs concomitantly with IL-2 production. When T cells were stimulated with sAbs neither CD69 nor CD25 were upregulated (Fig. 2.1b). In agreement with the impaired activation, crosslinking of the antibodies in solution also did not induce T-cell proliferation (Fig. 2.1c). Interestingly, we found that, conversely to mouse T cells where sAbs induce apoptosis (Wang et al, 2008), stimulation of human T cells with sAbs induces an anergic-like state, characterized by unresponsiveness to re-stimulation (data not shown).
Sustained signaling induced by iAbs
Antibodies against the TCR/CD3 complex can be bound to plastic (e.g. 96-well plates or plastic Petri dishes) or to microspheres of different size (Koike et al, 2003; Carpentier et al, 2009; Li et al, 2010). Under these conditions of stimulation, T-cell proliferation will be induced (Berg et al, 1998; Puente et al, 2000; Puente et al, 2006). In my work, I decided to take advantage of antibodies immobilized on SuperAvidin-coated microbeads with the diameter of approximately 10 µm. This size was selected to mimic the dimensions of APCs and to prevent endocytosis of the beads by T cells.
Next, I performed analysis of the signaling kinetics triggered by antibodies immobilized on microbeads (Fig. 2.2). In stark contrast to sAbs, iAbs stimulation triggered sustained phosphorylation of proximal and distal signaling molecules for up to 12 h (Fig. 2.2a and 2.2b). It has been previously shown that sustained signaling corresponds with T-cell activation and differentiation in mouse CD8+ T cells (Berg et al, 1998; Wang et al, 2008). Therefore, I tested whether iAbs also induce activation/proliferation of human T cells. As presented in figure 2.2c, microbeads were able to induce strong expression of both the
30
Figure 2.1. sAbs induce transient signaling and T-cell unresponsiveness. Purified primary human T cellswere treated with CD3, CD3xCD28 or CD3xCD4xCD28 mAbs cross-linked in solution (sAbs) as indicated. (a) Samples were analyzed by Western blotting using the indicated Abs. For each condition one representative immunoblot of at least four independent experiments is shown. (b) 24 h after stimulation with CD3xCD28 sAbs, the expression of CD25 and CD69 was analyzed by flow cytometry. (c) T cells were labeled with CFSE and stimulated as indicated in (b). Proliferation was assessed after 72 h by analyzing CFSE content by flow cytometry (the figure was modified from Arndt et al, 2013).
31
Figure 2.2. iAbs trigger sustained signaling kinetics, T-cell activation and proliferation. Purified primaryhuman T cells were treated with CD3, CD3xCD28 or CD3xCD4xCD28 mAbs immobilized on microbeads (iAbs) as indicated. (a and b) Samples were analyzed by Western blotting using the indicated Abs. For each condition one representative immunoblot of at least four independent experiments is shown. (c) 24 h after stimulation with CD3xCD28 sAbs, the expression of CD25 and CD69 was analyzed by flow cytometry. (d) T cells were labeled with CFSE and stimulated as indicated in (c). Proliferation was assessed after 72 h by analyzing CFSE content by flow cytometry (the figure was modified from Arndt et al, 2013).
32
activation markers CD69 and CD25. Moreover, flow cytometric analysis of CFSE-labeled T cells clearly showed that iAbs stimulation triggers proliferation (Fig. 2.2d).
Collectively, the data show that sAbs stimulation correlates with transient signaling that leads to unresponsiveness, whereas iAbs stimulation corresponds with sustained signaling and productive T-cell responses, such as CD69 and CD25 upregulation and proliferation.
2.2. Analysis of feedback regulation in transient vs. sustained T-cell activation
The stimuli I employed activate the same receptor and trigger the same cellular pathways. Nevertheless, they induce different activation dynamics, thus resulting in different cellular responses. Therefore, I performed additional studies to shed light onto how the activation of the same receptor (i. e. the TCR) results in different signaling kinetics.
2.2.1. Transient signaling correlates with a strong activation of Src family kinases
I initially focused my attention on proximal TCR signaling events (see 1.3.2). To this end, I measured the phosphorylation of TCRζ upon sAbs vs. iAbs treatment (Fig. 2.3a). TCRζ immunoprecipitates were prepared from either unstimulated or stimulated T cells and probed with an anti-pan-phospho-tyrosine antibody (clone 4G10). As shown in figure 2.3a, transient signaling (sAbs) correlates with increased tyrosine phosphorylation of TCRζ, as suggested by the appearance of two phosphorylated bands running at 21 and 23 kDa. In stark contrast, I did not observe any significant change in TCRζ chain phosphorylation during sustained signaling (iAbs) (Fig. 2.3a). These data imply that the activity of SFKs, which are responsible for TCRζ phosphorylation (see 1.3.2), is likely differentially regulated upon transient vs. sustained activation. Therefore, to further test this hypothesis, I analyzed the activity of Lck and Fyn. I focused on the fractions of Lck and Fyn which are associated with the TCR/CD3 complex and hence are directly involved in signaling. To this aim, I performed immunoprecipitations of TCRζ under mild detergent conditions to pull-down intact TCRζ chains associated with effector molecules. Subsequently, taking advantage of the phospho-Src-Y416-specific antibody, which recognizes phosphorylation of the activatory tyrosines Y394
and Y416 in Lck and Fyn, respectively, I tested TCRζ immunoprecipitates for the presence of active SFKs. The results presented in figure 2.3b show that, in contrast to sustained signaling, the amount of active Lck and Fyn associated with the TCR significantly increases during transient activation. These data demonstrate that SFKs activity is strongly enhanced during transient signaling. In agreement with these findings, I observed that the global tyrosine phosphorylation pattern is strongly induced during transient activation (Fig. 2.3c). Conversely, tyrosine phosphorylation does not appear to be as strongly induced during sustained as during transient signaling (Fig. 2.3c).
33
Figure 2.3. Transient signaling correlates with a strong activation of Src family kinases. Purified humanT cells were treated with either soluble (sAbs) or immobilized (iAbs) CD3xCD28 mAbs for the indicated time periods. TCRζ immunoprecipitates (a and b) or total cell lysates (c) were prepared and analyzed by Western blotting using the indicated Abs. In (b), anti-TCRζ-coated agarose beads were incubated in lysate buffer without cells as a control (Ctrl). One representative immunoblot of at least three independent experiments is shown. The phosphorylation of TCRζ (a) or Fyn and Lck (b) was quantified using the 1D ImageQuant software and the values were normalized to the corresponding total TCRζ or Lck and Fyn signal, respectivelyl. Data represent the mean of phosphorylation levels shown as arbitrary units ± SEM of at least three independent experiments. Asterisk in (b) indicates the antibody heavy chain (figures 2.3a and b were modified from Poltorak et al, 2013).
34
In summary, these data suggest that a quantitative difference in the activation of Src family kinases distinguishes transient from sustained activation.
2.2.2. Activation of negative regulators during transient TCR-mediated signaling
The enhanced activation of Lck and Fyn implies that also the activation of downstream signaling molecules is augmented under conditions of stimulation inducing transient signaling. This hypothesis is further supported by the enhanced tyrosine phosphorylation pattern (Fig. 2.3c). Indeed, I found that the phosphorylation of ZAP70, LAT, and PLCγ1 (Fig. 2.1) is enhanced upon sAbs stimulation. I hypothesized that, in addition to positive regulators, also the activation of negative regulators of TCR signaling may be strongly augmented during transient signaling. If this holds true, an imbalance in the equilibrium between positive and negative regulators may explain why signaling is rapidly terminated upon transient stimulation.
It is known that some negative regulators depend on tyrosine phosphorylation for their activation. One of these is cCbl, an E3 ubiquitin ligase involved in the ubiquitinylation and degradation of TCRζ and ZAP70 (see 1.4.1). I analyzed cCbl phosphorylation under conditions of stimulation inducing either transient or sustained signaling. The results presented in figure 2.4 clearly show that stimulation with sAbs greatly increases cCbl phosphorylation above the basal level, whereas treatment with iAbs does not. In agreement with these findings, we found that ZAP70 expression was reduced upon sAbs treatment, likely indicating degradation (Poltorak et al, 2013). This hypothesis is further supported by the observation that anti-ZAP70 immunoblots revealed a particular pattern of ZAP70 migration on SDS-PAGE gel corresponding with ubiquitinylation (Wang et al, 2008). Thus, activation of cCbl and the subsequent degradation of signaling molecules such as ZAP70 upon sAbs stimulation may lead to a dampening of the TCR-mediated signaling and hence may contribute to the observed transient activation under this condition of stimulation (Poltorak et al, 2013).
Next, I tested whether other negative regulators are also activated during sAbs stimulation. Another inhibitory molecule activated by tyrosine phosphorylation is Dok2 (see 1.4.1). As described in 1.4.1, Dok2 can interfere with TCR signaling at several levels. Therefore, it was important to test whether Dok2 is also phosphorylated/activated under sAbs treatment. As expected, Dok2 phosphorylation was strongly enhanced upon transient stimulation (Fig. 2.4). In contrast, in iAbs-treated T cells, Dok2 phosphorylation was only slightly increased above the basal level (Fig. 2.4).
35
Figure 2.4. Activation of negative regulators during transient TCR-mediated signaling. Purified naïve CD4+human T cells were left untreated or treated with either soluble (sAbs) or immobilized (iAbs) mAbs for the specified time periods. Samples were analyzed by Western blotting using the indicated Abs. The phosphorylation of cCbl and Dok2 was quantified using the 1D ImageQuant software and the values were normalized to the corresponding β-actin signal. Data represent the mean of the phosphorylation levels shown as arbitrary units ± SEM of three independent experiments. One representative experiment of four is shown.
